The main aim of this proposal is to advance our understanding of RNA stability, folding and dynamics. As the biological function of an RNA molecule mostly depends on its shape, we believe that achieving this aim will require the development of (a) new approaches to characterize the geometry and stability of RNA shapes, (b) new techniques to characterize intermediates that appear upon RNA folding, and (c) new methods to describe the map between RNA sequence space and structure space. The key to the success of this proposal lies in its collaborative integration within the NIH National Center for Biomedical Computing "Physics-based Simulation of Biological Structures", hosted at Stanford University, and lead by Prof. Russ Altman. The expertise of several Pis at the Stanford Center, including Profs Altman, Levitt and Pande, in the field of RNA dynamics as wells as the availability of many computational methods for structural biology in the SimTK library they have developed will prove invaluable for success. In particular, we will collaborate on the following specific aims: (1) Develop fast, accurate and analytic methods to measure geometric properties of RNA molecules, based on the successful methods we have originally developed for studying protein solvation. We are interested in particular in computing surface accessibility, and in correlating this geometric measure to experimental data such as hydroxyl radical footprinting. (2) Develop a new formalism for computing the electrostatic field around RNA. We propose to use a generalized Poisson-Boltzmann-Langevin equation to describe the electrostatic field generated by RNA in water, based on a description of the solvent as an assembly of freely orienting dipoles. (3) Use orthogonal normal modes in torsion and Cartesian space to characterize RNA dynamics. We are particularly interested in defining pathways between potential conformations of a RNA molecule using a combination of physical or elastic normal modes. (4) Describe the sequence space compatible with a RNA tertiary structure. We will apply the methods we have developed for protein sequence design to measure the ability of RNA to accept mutations without losing its structure. (5) Visualization of the geometric and physical properties of RNA structures. We will expand the ToRNADo RNA visualization platform developed at Stanford into an effective interface between theory and biology.